A Semi-parametric Estimation of Personalized Dose-response Function Using Instrumental Variables
Authors: Wei Luo, Yeying Zhu, Xuekui Zhang, Lin Lin
JMLR 2024 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Its effectiveness is illustrated by the simulation studies and the data analysis of ADNI-Do D study, where the causal relationship between depression and dementia is investigated. |
| Researcher Affiliation | Academia | Wei Luo EMAIL Center for Data Science Zhejiang University Hangzhou, P.R.China Yeying Zhu EMAIL Department of Statistics and Actuarial Science University of Waterloo Waterloo, ON N2L 3G1, Canada Xuekui Zhang EMAIL Department of Mathematics and Statistics University of Victoria Victoria, BC V8P 5C2, Canada Lin Lin EMAIL Department of Biostatistics and Bioinformatics Duke University Durham, NC 27710, USA |
| Pseudocode | Yes | Algorithm 1 Algorithm for the estimation of SPDRF Step 0. Calculate the initial value of e A using the lasso penalty in (19). Step 1. Given e A, calculate ˇΓ = bβY bβT e A. Let eΓ = ˇΓ, but, for each i I, modify eΓi to be ... Step 2. Given eΓ, calculate e A = (bβT T bβT) 1 bβT T(bβY eΓ). Step 3. Iterate between Step 1 and Step 2 until a convergence threshold is met. The most updated results are bΓ and b A, respectively. |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described. It only mentions the license for the paper itself and does not include an explicit statement of code release or a link to a code repository. |
| Open Datasets | Yes | Data used in preparation of this article were obtained from the Alzheimer Disease Neuroimaging Initiative (ADNI) database (adni.loni.usc.edu). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report. A complete listing of ADNI investigators can be found at: http://adni.loni.usc.edu/wpcontent/uploads/how to apply/ADNI Acknowledgement List.pdf |
| Dataset Splits | Yes | Following Fan and Li (2001), a in Step 1 is fixed at 3.7, and λ is tuned by a five-fold cross validation. ... and we recommend using a grid point search with five-fold cross validation to tune h and b. |
| Hardware Specification | No | The paper does not provide specific hardware details (like GPU/CPU models or memory) used for running its experiments. It only describes simulation studies and real data analysis without mentioning the computational environment. |
| Software Dependencies | No | The paper mentions statistical methods and penalties like SCAD (Fan and Li, 2001), lasso (Tibshirani, 1996), SIR (Li, 1991), SAVE (Cook and Weisberg, 1991), and lasso SIR (Lin et al., 2019), but it does not specify any version numbers for general software dependencies or programming languages used for implementation. |
| Experiment Setup | Yes | To implement the proposed method, we set τ = 1 in (27), and, depending on whether a symmetric data pattern exists in the data, we use SIR (Li, 1991) or SAVE (Cook and Weisberg, 1991) to estimate ST |X and SE(Y |X). In real data analysis, such pattern can be detected or excluded in the stage of exploratory data analysis. For the reliability of the proposed method, we first use (32) to check Assumption 2, where the bandwidths of Wτ are tuned by the cross-validation procedure mentioned in Section 6. ... For example, if the estimators of SE(Y |X) and ST |X are asymptotically normal, which, by Delta method and the strong oracle property in Theorem 2, imply the asymptotic normality of the nonzero rows of bΓ, then the ladle estimator (Luo and Li, 2016) is applicable. We also recommend using BIC (Zhu et al., 2006) and PAE (Luo and Li, 2021) if asymptotic normality is not guaranteed in the estimation of SE(Y |X) or ST |X. |